Windows Server Monitoring: The Half Perfmon Misses
Windows server monitoring needs two halves: the in-box Microsoft tools for resources, and external checks on every port the server promises to answer.
7 posts tagged "Tcp".
Windows server monitoring needs two halves: the in-box Microsoft tools for resources, and external checks on every port the server promises to answer.
A rabbitmq health check proves one thing at a time. Here is the ladder from ping to alarms, the HTTP endpoints, and the outage every ping survives.
A mysql health check built on mysqladmin ping reports success with the wrong password, and a socket probe matches the temporary server on first boot.
A kafka health check answers one of two questions: can the cluster serve traffic, or is this client still working. Here is the right probe for each.
A redis health check built on PING alone misses eviction, replication and failed saves. Here is the compose config and the probe that catch them.
A postgres health check needs three layers: pg_isready in the container, a real query behind your health route, and a port check from outside.